Racial and Gender Bias in Teacher Feedback

• In Event: Examining the Role of Racial-Ethnic Biases in Teacher Perceptions and Decision Making

Abstract

Teacher feedback plays a key role in student mathematics learning (e.g., Gentrup et al., 2020; Hu et al., 2021). This is particularly true for students who need additional guidance. Their errors provide valuable insights into their thinking processes, which allows teachers to direct students toward deeper understanding. Because feedback is part of teachers’ daily work, understanding potential gender and racial biases in teachers’ comments provides insights into the inequity in mathematics classrooms. In this study, we conducted an experiment to explore the potential systemic bias in teachers’ feedback to students’ solutions. We collected data from a national sample of U.S. elementary and middle school teachers (N = 510) who did not know that the students’ race and gender were randomly assigned to the same student solutions. We asked the teachers to provide feedback to improve the understanding of the students whose solutions were presented.

The majority of our sample was female (81.6%) and White (66.7%). Seventy-eight percent of the sample held a credential in teaching multiple subjects, whereas almost 16% held a single-subject credential in teaching mathematics. Each teacher was asked to provide feedback on four student solutions that were obtained from real students in the middle grades who had solved two mathematics problems drawn from the National Assessment of Educational Progress. These student solutions were incorrect or partially correct so that teachers would have room to provide feedback to improve the students’ understanding.

Teachers’ responses were coded according to a set of rubrics developed from prior work on capturing teachers’ analysis of student solutions (i.e., focusing on the strengths and weakness of the solution), their focus on the feedback (e.g., how students solved the problem, i.e., their process or only the final answer), and their perception of an individual student’s capacity based on the student’s work (e.g., the effort students put into the work).

The data were blinded and then coded by two researchers who did not know the actual purpose of the study (i.e., analyzing the gender and racial differences in teachers’ feedback). The interrater agreement between the raters ranged from moderate to substantial (kappa statistics were .71, .59, and .55 for each rubric, respectively). We then analyzed the data by using nominal logistic regressions for each rubric. The analysis predicted each outcome variable as a function of students’ presented gender and race, along with item and item order as fixed effects.

Our findings indicated that teachers were more likely to praise the effort students put into their work when they saw the image a White girl versus a Black boy (an effect size of .03, p = .002). We also found, albeit marginally significantly, that White boys received more process-oriented feedback than did Black boys (an effect size of .05, p = .079). Our findings suggest that students’ gender and race were linked to the kinds of feedback teachers gave in a way that limited opportunities for Black students to improve their mathematics understanding.

Author

• Yasemin Copur-Gencturk, University of Southern California